Systems, and methods are provided for predicting a cache policy based on application input data. Inputs provided to an application and corresponding to a usage pattern of the application can be received. The inputs can be used with a predictive model to determine a cache policy corresponding to a datastore. The cache policy can include output data to be provided via in the datastore and subsequently provided to a computing device in a just-in-time manner. The predictive model can be trained to output the cache policy based on input data received from a usage point, a provider point, or a datastore configuration.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method comprising: receiving by a server including a first datastore, data characterizing a pattern of sequential inputs provided to an application monitoring a cluster of oil and gas refinery equipment, the application configured on a computing device including a data processor, the computing device coupled to the server, the data wherein the pattern of sequential inputs corresponds to a usage pattern of the application and is provided with respect to an objective to be performed using the application in an oil and gas computing environment coupled to the cluster of oil and gas refinery equipment; determining, using the pattern of sequential inputs and a predictive model, a cache policy corresponding to the first datastore and including output data to be provided via the datastore, the cache policy predicted via the predictive model, wherein the predictive model is trained in a machine learning process to output the cache policy; executing the cache policy at the first datastore, the execution causing the first datastore to provide the output data to the application based on the pattern of sequential inputs provided to the application; providing, by the application, an output including the output data, wherein the output is provided based on executing the cache policy; and executing, by the application, at least a portion of the objective using the output.
This invention relates to optimizing data caching in oil and gas refinery monitoring applications. The problem addressed is inefficient data retrieval in applications that monitor clusters of refinery equipment, leading to delays in decision-making and operational inefficiencies. The solution involves a server with a datastore that receives sequential input patterns from a monitoring application running on a computing device. These inputs correspond to usage patterns of the application, which is designed to perform specific objectives in an oil and gas computing environment linked to refinery equipment. A predictive model, trained via machine learning, analyzes these patterns to determine an optimal cache policy for the datastore. This policy specifies which output data should be cached and retrieved to improve application performance. The server executes the cache policy, causing the datastore to provide the necessary data to the application, which then generates outputs based on this cached data. The application uses these outputs to execute at least part of its intended objective, such as monitoring or controlling refinery operations. The predictive model ensures that the cache policy dynamically adapts to usage patterns, reducing latency and improving efficiency in data-intensive refinery monitoring tasks.
2. The method of claim 1 , wherein the objective includes at least one of an operational objective associated with operating the cluster of oil and gas refinery equipment coupled to the oil and gas computing environment, a diagnostic objective diagnosing the oil and gas refinery equipment coupled to the oil and gas computing environment, analytical objective analyzing the oil and gas refinery equipment coupled to the oil and gas computing environment, or a search objective associated with searching for information related to the oil and gas refinery equipment coupled to the oil and gas computing environment.
This invention relates to methods for managing and optimizing oil and gas refinery operations using a computing environment. The method involves defining an objective for interacting with a cluster of refinery equipment, where the objective can include operational tasks, diagnostics, analysis, or information retrieval. Operational objectives involve controlling or monitoring the refinery equipment to ensure efficient and safe operation. Diagnostic objectives focus on identifying and troubleshooting issues within the equipment. Analytical objectives involve assessing performance data to optimize processes or predict maintenance needs. Search objectives enable users to retrieve relevant information about the equipment, such as historical data, specifications, or maintenance records. The computing environment processes these objectives by accessing and analyzing data from the refinery equipment, providing actionable insights or automated control adjustments. This approach enhances operational efficiency, reduces downtime, and improves decision-making in oil and gas refinery management.
3. The method of claim 1 , wherein the computing device includes one of a usage point, a provider point, a second datastore, and a data source.
A system and method for managing data processing involves a computing device that interacts with various components to facilitate data operations. The computing device is configured to handle data from multiple sources, including a usage point, a provider point, a second datastore, and a data source. The usage point represents an endpoint where data is consumed or utilized, while the provider point acts as a source of data or services. The second datastore serves as an additional storage location for data, distinct from a primary datastore, enabling redundancy or distributed processing. The data source provides raw or processed data to the computing device for further handling. The computing device processes, stores, or transmits data between these components, ensuring efficient data flow and management. This system addresses challenges in data integration, storage, and retrieval by providing a flexible architecture that accommodates different types of data endpoints and storage solutions. The method ensures seamless interaction between these elements, optimizing data accessibility and reliability.
4. The method of claim 3 , wherein the first datastore or the second datastore is associated with an application provider or a third party.
A system and method for managing data storage in a distributed computing environment addresses the challenge of efficiently organizing and accessing data across multiple storage locations. The invention involves a method for storing and retrieving data using at least two datastores, where each datastore is associated with either an application provider or a third-party entity. The method includes determining a storage location for data based on predefined criteria, such as data type, access frequency, or security requirements. The data is then stored in the appropriate datastore, and metadata is generated to track the storage location and facilitate retrieval. When data is requested, the system uses the metadata to locate and retrieve the data from the correct datastore. The method ensures data is stored in the most suitable location while maintaining accessibility and security. The association of datastores with either the application provider or a third party allows for flexible storage management, enabling cost optimization, performance improvements, and compliance with data governance policies. This approach is particularly useful in cloud computing environments where data may be distributed across multiple providers or services.
5. The method of claim 1 , wherein the cache policy includes an expiration parameter specifying an expiration period for the output data to be available in the first datastore.
This invention relates to data caching systems, specifically methods for managing cached data in distributed computing environments. The problem addressed is inefficient data retrieval in systems where frequently accessed data is stored in multiple datastores, leading to inconsistent or stale data being served to users. The invention provides a solution by implementing a cache policy that includes an expiration parameter to control how long output data remains available in a primary datastore before being refreshed or invalidated. The method involves storing output data in a first datastore, where the data is accessible to clients or other systems. A cache policy is applied to this data, with the policy including an expiration parameter that defines a specific time period during which the data remains valid in the first datastore. Once this period elapses, the data is either refreshed from a source or removed from the cache. The expiration parameter ensures that data remains current and reduces the risk of serving outdated information. The system may also include a second datastore that stores the same or related data, with synchronization mechanisms to maintain consistency between the two datastores. The expiration parameter can be dynamically adjusted based on factors such as data access patterns, system load, or user requirements, optimizing performance and resource usage. This approach improves data consistency, reduces redundant storage, and enhances system reliability in distributed environments.
6. The method of claim 5 , further comprising removing the output data from the first datastore at an end of the expiration period or based on receiving data characterizing a second pattern of sequential inputs provided to the application.
This invention relates to data management in computing systems, specifically for handling output data generated by applications. The problem addressed is the need to efficiently manage and remove output data from a datastore based on predefined conditions, such as an expiration period or changes in user input patterns. The method involves monitoring an application that generates output data, which is stored in a first datastore. The system tracks the output data and determines whether it should be retained or removed based on two possible conditions. First, the output data may be removed at the end of a predefined expiration period. Second, the system may analyze a second pattern of sequential inputs provided to the application and, based on this analysis, decide to remove the output data if the input pattern changes significantly. This ensures that outdated or irrelevant data is automatically purged, optimizing storage and improving system performance. The method may also involve additional steps, such as transferring the output data to a second datastore before removal, depending on the specific implementation. The overall goal is to automate data lifecycle management, reducing manual intervention and ensuring data relevance.
7. The method of claim 1 , wherein the output data is formatted based on the first datastore, the application, or a named user of the application.
A system and method for processing and formatting data from multiple sources involves retrieving data from a first datastore and a second datastore, where the second datastore contains metadata associated with the first datastore. The system generates output data by combining the retrieved data from both datastores, where the output data is formatted based on the structure or characteristics of the first datastore, the specific application requesting the data, or the identity of the user accessing the application. The method ensures that the output data is tailored to the needs of the application or user, improving usability and compatibility. The system may also include a user interface for displaying the formatted output data, allowing users to interact with the combined information efficiently. This approach enhances data integration and presentation, addressing challenges in managing and utilizing heterogeneous data sources.
8. The method of claim 1 , wherein the predictive model is trained in the machine learning process configured to generate the predictive model based on the pattern of sequential inputs collected from a usage point within the oil and gas computing environment, a provider point within the oil and gas computing environment, or a data source within the oil and gas computing environment.
This invention relates to machine learning in the oil and gas industry, specifically improving predictive models by training them on sequential input patterns from different points within the computing environment. The core problem addressed is the need for accurate, context-aware predictions in oil and gas operations, where data sources vary widely in structure and relevance. The method involves training a predictive model using sequential input patterns collected from at least one of three key sources: usage points, provider points, or external data sources within the oil and gas computing environment. Usage points refer to locations where operational data is generated, such as sensors or monitoring systems. Provider points are systems or services that supply data, such as cloud platforms or third-party analytics tools. External data sources may include industry databases, weather data, or market trends. The training process leverages these sequential inputs to identify patterns that improve the model's predictive accuracy. For example, sensor data from a drilling operation (usage point) may be combined with real-time weather data (external source) to forecast equipment failures. Similarly, provider point data, such as cloud-based analytics outputs, can be integrated to refine predictions. By incorporating diverse, time-sequenced data streams, the model adapts to dynamic conditions in oil and gas operations, enhancing decision-making for maintenance, production optimization, and risk management. The approach ensures the model remains relevant as new data sources or operational changes occur.
9. The method of claim 8 , wherein the machine learning process is configured to generate a plurality of predictive models based on a predetermined schedule, wherein each predictive model of the plurality of predictive models includes at least one new or updated cache policy.
This invention relates to machine learning-based systems for optimizing cache policies in computing environments. The problem addressed is the inefficiency of static or manually configured cache policies, which fail to adapt to dynamic workloads and changing data access patterns, leading to suboptimal performance and resource utilization. The invention describes a method for dynamically generating and updating cache policies using machine learning. A machine learning process is trained on historical data to predict future data access patterns and system performance metrics. The process generates multiple predictive models on a predetermined schedule, such as hourly, daily, or weekly, to ensure continuous adaptation to evolving workloads. Each predictive model includes at least one new or updated cache policy, which may involve adjustments to cache size, eviction strategies, or prefetching mechanisms. The generated policies are then deployed to optimize cache performance, reducing latency and improving resource efficiency. The machine learning process may incorporate various techniques, such as reinforcement learning or supervised learning, to refine the predictive models over time. The system may also validate the generated policies using simulation or real-world testing before deployment. By automating the policy generation process, the invention eliminates the need for manual tuning, ensuring that cache performance remains optimal under varying conditions. This approach is particularly useful in high-performance computing, cloud environments, and distributed systems where workloads are dynamic and unpredictable.
10. The method of claim 9 , wherein the pre-determined schedule specifies collecting the data characterizing the pattern of sequential inputs provided to the application continuously, every hour, every day, every week, every month, or during a user-defined time-period.
This invention relates to a method for collecting and analyzing data characterizing the pattern of sequential inputs provided to an application. The method addresses the challenge of efficiently gathering and processing input data to understand user behavior, system performance, or other relevant metrics over time. The method involves collecting data at predetermined intervals, which can be continuous, hourly, daily, weekly, monthly, or during a user-defined time period. This allows for flexible monitoring and analysis of input patterns, enabling insights into usage trends, performance bottlenecks, or other critical factors. The collected data can be used to optimize application functionality, improve user experience, or detect anomalies. The method ensures that data is gathered systematically, providing a comprehensive view of input sequences over the specified time frames. By adjusting the collection frequency, users can balance between granularity and resource efficiency, depending on the specific requirements of the application or analysis. This approach enhances decision-making by providing actionable insights derived from structured input data.
11. A system comprising: at least one computing device including an application monitoring a cluster of oil and gas refinery equipment, the application configured on the at least one computing device; a server coupled to the at least one computing device via a network, the server including a processor and a memory storing computer-readable instructions, a plurality of prediction models, and a first datastore, the processor configured to execute the computer-readable instructions, which when executed, cause the processor to perform operations comprising: receiving data characterizing a pattern of sequential inputs provided to the application configured on the computing device and coupled to a datastore, the pattern of sequential inputs corresponding to a usage pattern of the application and provided with respect to an objective to be performed using the application in an oil and gas computing environment coupled to the cluster of oil and gas refinery equipment; determining, using the pattern of sequential inputs and a predictive model, a cache policy corresponding to the first datastore and including output data to be provided via the datastore, the cache policy predicted via the predictive model, wherein the predictive model is trained in a machine learning process to output the cache policy; executing the cache policy at the first datastore, the execution causing the first datastore to provide the output data to the application based on the pattern of sequential inputs provided to the application; providing, by the application, an output including the output data, based on executing the cache policy; and executing, by the application, at least a portion of the objective using the output.
The system optimizes data access in oil and gas refinery operations by dynamically adjusting cache policies based on application usage patterns. In this domain, refinery equipment generates large volumes of data, and inefficient data retrieval can lead to delays and operational inefficiencies. The system includes a computing device running an application that monitors refinery equipment, connected to a server via a network. The server stores prediction models and a datastore, executing instructions to analyze sequential inputs from the application, which reflect how users interact with the system to achieve specific objectives, such as optimizing refinery processes. A predictive model, trained via machine learning, processes these inputs to determine an optimal cache policy for the datastore, specifying which data should be cached to improve performance. The server then applies this policy, ensuring the datastore provides the most relevant data to the application, which uses it to execute tasks efficiently. This approach reduces latency and enhances decision-making in refinery operations by aligning data access with real-time usage patterns. The system improves operational efficiency by dynamically adapting to changing data demands in the oil and gas computing environment.
12. The system of claim 11 , wherein the objective includes at least one of an operational objective associated with operating the cluster of oil and gas refinery equipment coupled to the oil and gas computing environment, a diagnostic objective associated with diagnosing the oil and gas refinery equipment coupled to the oil and gas computing environment, an analytical objective associated with analyzing the oil and gas refinery equipment coupled to the oil and gas computing environment, or a search objective associated searching for information related to the oil and gas refinery equipment coupled to the oil and gas computing environment.
The system is designed for managing and optimizing oil and gas refinery operations through a computing environment. The system addresses challenges in monitoring, diagnosing, and analyzing refinery equipment, which are critical for maintaining efficiency, safety, and productivity in oil and gas processing. The system includes a computing environment that interfaces with a cluster of refinery equipment, enabling real-time data collection, processing, and decision-making. The system supports multiple objectives, including operational objectives to ensure smooth and efficient equipment functioning, diagnostic objectives to identify and resolve equipment issues, analytical objectives to assess performance and optimize processes, and search objectives to retrieve relevant information about the refinery equipment. These objectives are tailored to enhance operational reliability, reduce downtime, and improve overall refinery performance. The system integrates data from various sources, such as sensors and control systems, to provide actionable insights and support decision-making in refinery management.
13. The system of claim 12 , wherein the computing device includes one of a usage point, a provider point, a second datastore, and a data source.
A system for managing and processing data in an energy management or industrial control environment addresses the challenge of efficiently collecting, storing, and analyzing data from multiple sources. The system includes a computing device that interfaces with various components, such as a usage point (e.g., a meter or sensor), a provider point (e.g., an energy supplier or control device), a second datastore (e.g., a database for historical or reference data), and a data source (e.g., an external system or sensor network). The computing device is configured to receive, process, and transmit data between these components, enabling real-time monitoring, control, and decision-making. The system may also include a primary datastore for storing processed data, ensuring data integrity and accessibility. The computing device can execute algorithms to analyze data, detect anomalies, optimize resource allocation, or generate reports. This system improves operational efficiency by centralizing data management and enabling seamless integration between different data sources and storage systems. The invention is particularly useful in applications requiring precise data tracking, such as energy distribution, industrial automation, or smart grid management.
14. The system of claim 13 , wherein the first datastore or the second datastore is associated with an application provider or a third party.
A system for managing data storage and access in a distributed computing environment addresses the challenge of efficiently organizing and retrieving data across multiple storage locations. The system includes a first datastore and a second datastore, each capable of storing data objects. A data management module processes requests to store or retrieve data objects, determining which datastore should handle the request based on predefined criteria such as data type, access frequency, or storage cost. The system ensures data consistency and availability by synchronizing data between the datastores when necessary. The first or second datastore may be associated with an application provider or a third-party service, allowing for flexible deployment models where data can be stored either within the provider's infrastructure or externally. This modular approach enables scalability and cost optimization by leveraging different storage solutions based on operational needs. The system also supports metadata tagging and indexing to improve searchability and retrieval performance. By dynamically routing data requests to the appropriate datastore, the system enhances efficiency and reduces latency in data access operations.
15. The system of claim 11 , wherein the cache policy includes an expiration parameter specifying an expiration period for the output data to be available in the first datastore.
A system for managing data storage and retrieval in a distributed computing environment addresses the challenge of efficiently caching and accessing frequently used data while minimizing latency and resource consumption. The system includes a first datastore for storing output data generated by a processing module, a second datastore for storing input data, and a cache policy that governs how data is stored and retrieved. The cache policy defines rules for determining when data should be stored in the first datastore and how long it should remain available. The system dynamically adjusts storage and retrieval operations based on the cache policy to optimize performance. The system also includes a monitoring module that tracks data access patterns and updates the cache policy accordingly. The expiration parameter within the cache policy specifies a time period after which the output data in the first datastore is no longer considered valid and may be removed or refreshed. This ensures that stale data does not persist, maintaining data accuracy while balancing storage efficiency. The system may also include a synchronization mechanism to ensure consistency between the first and second datastores. The overall approach improves data availability and reduces redundant processing by intelligently managing cached data.
16. The system of claim 15 , further comprising removing the output data from the first datastore at an end of the expiration period or based on receiving data provided characterizing a second pattern of sequential inputs provided to the application.
A system for managing data in a computing environment addresses the challenge of efficiently handling temporary or time-sensitive data generated by applications. The system includes a first datastore configured to store output data generated by an application in response to sequential inputs, such as user interactions or automated processes. The system also includes a second datastore that stores metadata associated with the output data, including an expiration period for the data. A processing module monitors the first datastore and the second datastore to determine when the expiration period for the output data has ended. Upon detecting the expiration, the processing module removes the output data from the first datastore. Additionally, the system can remove the output data based on detecting a second pattern of sequential inputs to the application, which may indicate a change in usage or a need to purge the data. This ensures that the system dynamically adapts to usage patterns while maintaining data integrity and storage efficiency. The system may also include a user interface for configuring the expiration period or defining input patterns that trigger data removal.
17. The system of claim 11 , wherein the output data is formatted based on the first datastore, the application, or a specific named user of the application.
This invention relates to a data processing system that formats output data based on the source datastore, the application, or a specific user. The system retrieves data from a first datastore and processes it to generate output data. The output data is then formatted according to predefined rules associated with the first datastore, the application using the data, or a specific named user of the application. The formatting may include adjusting data structure, presentation, or metadata to ensure compatibility or usability. The system may also include a second datastore for storing additional data or metadata related to the formatting rules. The formatting process ensures that the output data is optimized for the intended use case, whether for display, further processing, or integration with other systems. The system dynamically applies formatting rules to maintain consistency and relevance across different contexts. This approach improves data usability and reduces the need for manual adjustments, enhancing efficiency in data-driven applications.
18. The system of claim 11 , wherein the predictive model is trained in the machine learning process configured to generate the predictive model based on the pattern of sequential inputs collected from a usage point within the oil and gas computing environment, a provider point within the oil and gas computing environment, or a data source within the oil and gas computing environment.
This invention relates to a machine learning-based predictive modeling system for oil and gas computing environments. The system addresses the challenge of optimizing operations by predicting outcomes based on sequential data patterns from various sources within the oil and gas industry. The predictive model is trained using machine learning techniques to analyze sequential inputs collected from usage points, provider points, or external data sources within the oil and gas computing environment. Usage points may include operational sensors or monitoring systems that track equipment performance, while provider points could involve supply chain or service provider data. The model learns from these patterns to forecast future states, such as equipment failures, production trends, or maintenance needs. By integrating data from multiple sources, the system enhances accuracy and adaptability in dynamic oil and gas operations. The training process involves feeding the model historical and real-time sequential data to identify correlations and trends, enabling proactive decision-making. This approach improves efficiency, reduces downtime, and optimizes resource allocation in the oil and gas sector. The system is designed to be scalable and adaptable to different operational contexts within the industry.
19. The system of claim 18 , wherein the machine learning process is configured to generate a plurality of predictive models based on a predetermined schedule, wherein each predictive model of the plurality of predictive models includes at least one new or updated cache policy.
A system for optimizing cache performance in computing environments addresses the challenge of efficiently managing data storage and retrieval in systems with limited cache resources. The system employs machine learning to dynamically adjust cache policies based on usage patterns and performance metrics. A machine learning process generates multiple predictive models on a predetermined schedule, such as hourly, daily, or weekly. Each model produces at least one new or updated cache policy, which may include rules for data eviction, prioritization, or allocation within the cache. The system evaluates these policies using historical and real-time data to select the most effective ones, improving cache hit rates and reducing latency. The machine learning process may incorporate various algorithms, such as reinforcement learning or supervised learning, to adapt to changing workloads and system conditions. The system also monitors cache performance metrics, such as access times and eviction rates, to refine the models over time. By continuously updating cache policies, the system ensures optimal performance without manual intervention, making it suitable for high-performance computing, databases, and distributed systems.
20. The system of claim 19 , wherein the pre-determined schedule specifies collecting the data characterizing the pattern of sequential inputs provided to the application continuously, every hour, every day, every week, every month, or during a user-defined time-period.
This invention relates to a system for monitoring and analyzing user input patterns in software applications. The system addresses the challenge of understanding how users interact with applications over time, which is critical for improving user experience, identifying inefficiencies, and detecting anomalies. The system collects data characterizing the sequence of inputs provided by users to an application, such as keystrokes, mouse movements, or touch gestures, and analyzes this data to identify patterns. These patterns can reveal usage trends, common workflows, or potential usability issues. The system includes a data collection module that gathers input data from the application, a storage module that stores the collected data, and an analysis module that processes the data to extract meaningful patterns. The system can operate on a predetermined schedule, collecting data continuously or at specified intervals, such as hourly, daily, weekly, monthly, or during a user-defined time period. This flexibility allows the system to adapt to different monitoring needs, whether for real-time analysis or periodic reviews. The system may also include a reporting module to present the analyzed patterns in a user-friendly format, such as visualizations or summaries, to aid in decision-making. By continuously or periodically monitoring input patterns, the system helps developers and analysts gain insights into user behavior and optimize application design.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
April 9, 2020
February 15, 2022
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.